Where Knowledge Begins? Generative Search, Information Literacy, and the Problem of Friction

Abstract This article explores the rise of generative AI and large language model (LLM) tools for internet search and their potential impact on student information seeking behavior. Reviewing what is known about best search practices, the essay identifies a schism between AI companies’ vision of search and information science's understanding of humane search environments. It argues that generative search tools fail to create ethical spaces for search, leading to dangerous territories for knowledge production. Specifically, the article discusses “friction” as a critical concept where the field of information science and AI development philosophies diverge. Whereas information science views friction as a valuable and often necessary component to search, AI companies view friction as a problem to eliminate. It is this mismatch between corporate AI philosophies and known best practices for search that, this essay argues, renders current LLM and generative AI search tools fundamentally incompatible with ethical processes of knowledge production.